When using AI, approach it with skepticism and ask probing questions to uncover its flaws.

Testing is essential for ensuring that AI systems behave as expected, not just in terms of performance but also in their underlying behaviors.

To successfully implement AI agents, start small with a well-known task, design for limited downside, leverage natural language for building, iterate with compassion, and gradually increase the agent's scope and responsibility.

An iterative approach is essential in working with AI; continuously debugging and refining prompts can lead to more reliable outcomes.

The ability to prompt AI effectively is crucial, as it determines the quality of the output, which is a reflection of agency and taste in AI development.

It's crucial to understand the complexities of working with AI systems, including the need to manage issues like hallucinations and model evaluations effectively.

When AI gives mediocre results, treating it as a teammate means providing feedback and coaching to improve its output.

Adversarial testing involves questioning AI outputs from multiple angles, as the AI will not tire or take offense, allowing for thorough exploration of its responses.

The challenge with agentic AI is balancing its capabilities with necessary guardrails to prevent misuse.